The Nonprofit Doing the AI Industry's Dirty Work

Calling archiving the web for researchers dirty work is a bit much.

Unless something has changed since I was there, the crawler didn't intentionally bypass any paywalls.

The crawler obeyed robots.txt, throttled itself when visiting slow sites to avoid overloading them and announced its user agent clearly with a URL explaining what it was and how to block it if desired.

Dupe: https://news.ycombinator.com/item?id=45810135